International Journal of Thermal Sciences 168 (2021) 107064

Contents lists available at ScienceDirect

International Journal of Thermal Sciences

journal homepage: www.elsevier.com/locate/ijts

Study on thermal performance of new finned heat exchange tube bundles in

cooling tower
Rui Zhao a, Shi Bu a, Xuhui Zhao a, Lin Zhang a, b, *, Weigang Xu a, Zhikang Yu a, Jiamei Fang a, b,
Yixiang Ji a, Yiyang Hu a, Bingguo Bao a, c
a
School of Mechanical Engineering and Rail Transit, Changzhou University, Changzhou, China
b
Jiangsu Changnuo Energy Environmental Protection Technology Co., Ltd, Changzhou, 213000,China
c
Jiangsu Seagull Cooling Tower Co., Ltd., China

A R T I C L E I N F O A B S T R A C T

Keywords: The way of heat exchange in the closed wet cooling tower is through the direct contact between the air in the
Closed wet cooling tower tower and the process water, so as to achieve the purpose of reducing the temperature of the process water. In
Energy efficiency order to improve the thermal performance of closed wet cooling tower and solve the problem of low heat transfer
Heat exchange tube
efficiency, a new elliptical finned tube of cooling tower is proposed, and a mathematical model is established.
Heat transfer
The process of heat and mass exchange in the tower is studied by numerical method. The effects of wind velocity
Numerical simulation
and spray density on the thermal performance of cooling towers under different structures are analyzed. The
results show that the whole velocity field is uniformly distributed, and there is an obvious flow around the
cylinder on both sides of the pipe wall, and with the increase of air velocity in the tower, the flow around the
cylinder is more obvious, compared with the traditional circular heat exchanger tube, the new elliptical finned
tube can significantly improve the thermal performance of the closed wet cooling tower, meet the needs of
practical engineering, and reduce the energy consumption of the cooling tower.

mechanical ventilation cooling towers under different heat loads and

adapt to various application environments. Previous studies have pro­
1. Introduction vided us with a lot of references, although many optimization methods
simplify the calculation and improve the applicability to a certain
The cooling tower has unique advantages in the treatment of high extent, the accuracy needs to be further verified. Poppe method is a new
temperature cooling water [1]. It is of great significance to fully analyze method for calculating the thermal performance of cooling towers on the
the thermal characteristics of the cooling tower, improve the cooling basis of Merkel theoretical analysis. Zheng W Y [12] derived the heat
efficiency of the cooling tower and reduce the energy consumption in and mass transfer control equation for unsaturated or supersaturated
the process of cooling water operation [2–6]. closed wet cooling tower based on Poppe method. Two analytical
By optimizing the calculation process and method, the heat and mass models are proposed, and the corresponding calculation programs are
transfer process of the cooling tower are theoretically analyzed. written based on the piecewise method. Through the establishment of
Sutherland [7] proposed a new differential analytical formula, and the numerical model and prediction scheme, the process of heat and mass
error between the results obtained by this formula and the experimental exchange in the tower is further explored. Zhou Y et al. [13] proposed a
results is less than 10%. Ren C Q [8] according to the heat exchange new scheme for predicting the thermal performance of cooling towers
process of heat exchanger, a new method (model) suitable for thermal based on unit or non-unit Lewis factors. In this scheme, the outlet water
calculation of cooling tower is put forward, which makes the calculation temperature is predicted based on the basic control equation of heat
process more convenient. Cui H J et al. [9] established a model exchange in a closed cooling tower. The predicted results are compared
considering the rise and fall of droplets, including mass, energy and with the numerical results to verify the effectiveness of the analytical
momentum, and discussed the displacement, velocity and temperature scheme. Xiaocui Xie et al. [14] established a two-dimensional model of
distribution of spray droplets in detail. Singh K [10,11] proposed an countercurrent closed cooling tower, and studied the influence of
inverse feedback model to obtain the thermal performance of

* Corresponding author. School of Mechanical Engineering and Rail Transit, Changzhou University, Changzhou, China.
E-mail address: z3281315@yeah.net (L. Zhang).

https://doi.org/10.1016/j.ijthermalsci.2021.107064
Received 27 January 2021; Received in revised form 24 April 2021; Accepted 10 May 2021
Available online 19 May 2021
1290-0729/© 2021 Elsevier Masson SAS. All rights reserved.
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Nomenclature Greek letters

α Heat transfer coefficient (w•m-2•s-1)
Cp specific heat capacity; (J/kg⋅k) β Mass transfer coefficient (kg•m-2•s-1)
A Heat or mass transfer area; (m2) η Cooling efficiency (%)
K thermal conductivity; (W/m⋅K) i Air enthalpy (kJ/kg)
m Mass; (kg)
Q heat exchange capacity; (W) Subscripts
t temperature; (◦ C) a Air
T Process water temperature; (◦ C) in Inlet
P Water vapor partial pressure; (pa) lh latent heat
i’ Saturated air enthalpy; (kJ/kg) masw Mean value of saturated wet air
D0 Tube diameter; (mm) out Outlet
γ0 Heat of vaporization of water at 0 ◦ C; (KJ/kg) s Spray water
u Uncertainty; (%) sw Saturated water
G Air mass flow; (kg/m2s) w Process water
L Length of heat exchange tube; (mm) wb Wet bulb
Nt Single tube bundle length; (mm) v Vapor
Wt Tube bundle width; (mm)

structure on the heat transfer characteristics of the cooling tower by new elliptical finned tube of the closed wet cooling tower is proposed,
simulation. The influence of operation/structure parameters on the the heat transfer mechanism of the closed wet cooling tower is analyzed,
cooling characteristics of cooling tower is summarized and the influ­ and the mathematical model of the closed wet cooling tower is estab­
ential trend of different grid number and turbulence model on the lished. The influence of the structure of the heat exchanger tube on the
calculation results is verified. Although the numerical simulation thermal performance of closed wet cooling tower is studied by numer­
scheme is becoming more and more perfect and the model is more and ical calculation, so as to achieve the purpose of strengthening the ther­
more accurate, there are still some errors compared with the actual mal performance of closed wet cooling tower.
operating results of the cooling tower. Mansour M K et al. [15,16]
established a new thermodynamic performance analysis model, which is 2. Geometry
based on the enthalpy potential method and combined with the ther­
modynamic equation. The calculated results are consistent with the The tube structure of the closed wet cooling tower is shown in Fig. 1.
experimental results. Li X et al. [17] studied the influence of a crosswind For the tube, the length L is much larger than its width H, so the flow
on the design and characteristics of natural ventilation cooling towers, characteristic of the fluid in the Z direction is much smaller than that in
and studied three common cooling towers through experimental the X direction and Y direction, and the fluid flow in the Z direction can
research and simulation. According to the comparison of the influence of be ignored. And the closed wet cooling tower has a symmetrical geo­
crosswind on different types of natural ventilation cooling and its action metric structure, so the calculation domain can be simplified to half of
mechanism in the existing literature, the accuracy and authenticity of the original calculation domain as the numerical calculation domain.
the results were verified. Most of these experiments study the accuracy Based on this study, a two-dimensional simplified model is selected to
of the new model and the influence of the external environment on the simulate and calculate the heat transfer efficiency of the closed wet
cooling tower, but there are few studies on the improvement of the in­ cooling tower.
ternal structure of the cooling tower. (Reviewer 2, Comment 1). The size of the heat exchange tube in the model is determined ac­
In this paper, in order to improve the thermal performance of the cording to the size of the cooling tower heat exchange tube in the actual
closed wet cooling tower, based on previous studies of the influence of engineering application. (Reviewer 2, Comment 4) The basic structural
different heat exchange tubes on the performance of cooling towers, the parameters of numerical calculation is the height of the numerical model

Fig. 1. Structural diagram.

2
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

is 297 mm, the length is 207 mm, the spray height is 20 mm, the heat 3.1.3. Component transport model
exchanger tube is replaced by a solid circle, the diameter of the tube is In order to simulate the mass transfer process between air and spray
20 mm, the distance between the tube and the tube is 42 mm, the water, the convection equation is used to predict the mass fraction in the
thickness of the tube wall is 2 mm, the diameter ratio of the elliptical whole calculation domain, when calculating the conservation of com­
heat exchanger tube is 0.6 mm, the length of the fin of the new elliptical ponents in Fluent. The equation is as follows:
finned tube is 3 mm and the width is 1 mm, the extension lines of the fins
∂ ( )
pass through the center of the ellipse, and the angle with the horizontal (ρYi ) + ∇ ⋅ ρ→
v Y j = − ∇ ⋅ J i + Ri + S i (6)
∂t
line is 45◦ , the specific parameters are shown in Table 1. The wall
temperature is constant, the spray height is still 20 mm, and the velocity Where: Ri is the rate of change of the rate of formation due to reaction in
of air in the tower is 2–4 m/s. According to the seasonal variation, the the process of mass transfer, Si is the rate of change of velocity produced
range of inlet temperature is 5–35 ◦ C, and the spray density is 0.02–0.07 by the motion of liquid droplets.
kg m− 2 s− 1. In the whole calculation domain, the heat exchanger tubes In the numerical calculation of this chapter, because the mass ex­
are arranged in 8 layers, with 4 tubes in each layer, arranged in the change between air and water is simulated, the component transport
following manner: staggered arrangement. Using the computing domain model without chemical reaction is selected, and in order to ensure the
grid divided by ANSYS ICEM, the whole grid is a highly orthogonal reliability of the calculation results, Eq. (6) can be simplified to
structured grid with a mass of more than 0.75, which is sufficient to meet Ref. [19]:
the requirements of grid quality in numerical calculation.
∂ ( ) →
(ρYi ) + ∇ ⋅ ρ→
v Yj = − ∇⋅ J i (7)
3. Computational method ∂t
→
3.1. Governing equation Where: J i represents diffusion density.
→ ∇T
3.1.1. Energy equation J i = − ρDi,m ∇Yi − DT,j (8)
T
The energy equation in the process of Fluent numerical calculation is
In the process of calculating the bulk mass coefficient caused by
shown in Eq. (1):
turbulence, Eq. (9) is selected in Fluent to calculate the specific change
( )
∂ ∑ → ( ) value:
(ρE) + ∇ ⋅ (→
v (ρE + p)) = ∇ ⋅ keff ∇T − hj Jj + τef f ⋅ →
v + Sh (1) ( )
∂t → μ ∇T
(9)
j
J i = − ρDi,m + t ∇Yi − DT,i
Sct T
Where: keff is thermal conductivity, Sh is chemical reaction heat.
Where: Sct is effective schmitt number of turbulence, μt is turbulent
3.1.2. Turbulence model viscosity.
Compared with Standard and RNG turbulence models, Realizable k-ε μt
turbulence model has the following advantages: Sct = (10)
ρDt

(1) The Realizable k-ε turbulence model contains equations for In summary, turbulent diffusion will affect the mass exchange be­
calculating the viscosity dependence of turbulence; tween air and droplets, which will change with the change of Schmidt
(2) A new method for calculating the dissipation rate ε is derived number. It can be concluded from Eq. (10) that there is laminar diffusion
from the calculation equation of vorticity [18]. in the process of turbulent diffusion, so there is no need to add new
[( ) ] properties to the laminar flow.
∂ ( ) ∂ μ ∂k
ρ kuj = μ+ i + Gkb − ρε (2)
∂xj ∂uj σ k ∂xj 3.1.4. DPM model
[( ) ] In the numerical calculation of Fluent, the inlet air is taken as the
∂ ( ) ∂ μi ∂ε ε2 continuous phase and the droplet as the discrete phase. The SIMPLE
ρ εuj = μ+ + ρC1 Sε − ρC2 √̅̅̅̅̅ (3)
∂xj ∂xj σε ∂xj k + vε algorithm is used to couple the temperature field and velocity field, and
the trajectory of the droplet in the whole calculation domain and its
Where: Gk is turbulent energy. influence on the heat transfer are analyzed and solved.
The eddy current viscosity in the turbulence model is calculated In the flow field of the whole computational domain, the control of
according to the following formula, but the k-ε Realizable turbulence the falling trajectory of simulated droplets is as follows:
model is calculated by Cμ and is not a fixed constant. ( )
( )
dup g ρp − ρg
k2 = FD ug − up + + Fx (11)
μ = ρC μ (4) dt ρp
ε

1 Where: FD represents the total force per unit mass of liquid droplets,
Cμ = * (5) FD(ug-up) represents drag force per unit mass droplet, Fx represents the
A0 + As kUε
droplet source term, which is used to describe other forces on the
Where: A0 is a fixed constant with a value of 4.3. droplet.
CD Rep
FD = (12)
24τP
Table 1
Basic structure parameters of the new elliptical finned tube (Reviewer 1, ρp dp2
Comment 2). τp = (13)
18μg
Parameter L H Tube Major Diameter Tube wall
spacing axis ratio thickness ⃒ ⃒
ρp dp ⃒up − ug ⃒
Numerical 3 1 42 25 0.6 2 Rep = (14)
μg
value (mm)

3
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Where: ug is fluid velocity, τp is droplet relaxation time, Rep is relative 3.4.2. Basic hypothesis
reynolds number of droplets, CD is drag coefficient, μg is hydrodynamic There are three different kinds of fluids in the closed wet cooling
viscosity. tower, which makes the heat transfer process in the tower very complex.
To sum up, the droplets are mainly affected by the diameter of Through certain simplification, not only the accuracy of calculation can
droplets and the velocity of inlet air in the calculation domain. In this be improved, but also the calculation time can be reduced. Therefore,
case, the influence of the droplet on the air flow during the falling the following assumptions are made in the study:
process is ignored, so the droplet diameter is usually set to a smaller
value or the DPM model is closed directly in the process of flow field (1) The heat exchange process of water-gas in the tower is regarded
calculation. as steady state.
(2) The heat transfer in the tower is a gas-liquid two-phase flow, and
both are regarded as incompressible fluids.
3.2. Boundary conditions (3) Once the spray droplet hits the wall of the computing domain, it
will be captured, which means that there is no liquid film on the
For the continuous phase (air), the velocity inlet and pressure outlet wall.
are used because the inlet velocity is constant and the direction is
perpendicular to the inlet boundary. The pressure outlet boundary 3.5. Performance evaluation
conditions are set by default, in which the inlet air is regarded as an
incompressible fluid. For the inlet and outlet of the closed wet cooling According to the operation parameters such as air temperature, air
tower, the turbulent boundary conditions are determined by turbulence velocity and spray water temperature at the outlet of the cooling tower
intensity and equivalent diameter. The inlet air velocity is set to 2–4 m/ obtained in the numerical calculation, the heat transfer unit number
s, the wet bulb temperature is set to 28 ◦ C, the inlet air temperature is set NTU and air enthalpy i2 are calculated from these parameters, and the
to 25–37 ◦ C, the spray water temperature is set to 35–41 ◦ C, and the outlet water temperature t, heat transfer coefficient β and mass transfer
spray density is set to 0.01–0.08 kg m− 2 s− 1. The specific boundary is coefficient α of the heat transfer tube can be calculated [20].
shown in Table 2. The number of heat transfer units of the cooling tower NTU is
calculated by Eq. (15):

3.3. Fluid properties T2 = t + (T1 − t)e− NTU

(15)

In the numerical calculation of the closed wet cooling tower, the Where: t is spray water temperature, T2 is influent temperature of heat
cooling efficiency varies with the moisture content of the inlet air and exchanger tube, NTU is number of heat transfer units, T1 is influent
the air density, and the cooling efficiency is related to the temperature temperature of heat exchanger tube.
difference between the air and the spray water, so it is necessary to set On the basis of the calculation of Eq. (15), the enthalpy value i2 of the
the physical parameters of the air. Set the air moisture content to air outlet is calculated, as shown in Eq. (16):
0.04075 kg/kg, the air density to 1.225 kg/m3, the water vapor mass ( )
i2 = i˝1 − i"t − i1 e− Mw (16)
fraction to 0007258.
Where: i2 is air outlet enthalpy, i1 is air inlet enthalpy, i1′′ is enthalpy of
3.4. Numerical solutions saturated air corresponding to process water temperature t, Mw is water
film cooling number.
3.4.1. Model setup As a result, the heat transfer coefficient of spray water to water film is
When using ANSYS Fluent to simulate the heat and mass exchange calculated, as shown in Eq. (17):
process between air and spray water in a closed wet cooling tower, the /
β = qp Cpp NTU A (17)
component transport model (SpeciesModel) is selected to simulate the
mass transfer process between air and spray water in the real environ­ Where: β is heat transfer coefficient of water film, qp is process water
ment, and the energy equation is used to simulate the heat transfer flow in heat exchanger tube, Cpp is specific heat of process water in heat
process between water and air in the real environment. At the same exchanger tube.
time, in order to ensure the accuracy of the flow field in the tower in the Based on the calculated results of Eq. (15) and Eq. (16), the outlet
process of numerical calculation, the second-order upwind scheme is water temperature of the heat exchanger tube of the cooling tower is
adopted and the SIMPLE algorithm is used to couple the pressure field obtained, which is calculated by Eq. (18):
and velocity field. In the process of solving the problem, the temperature
of the pipe wall is set to constant (constant wall temperature). In the t = T1 −
G(i1 − 1)(1 − e− Mw )
(18)
process of water-air heat transfer, the convergence residual values of the qp Cpp (1 − e− NTU )
energy equation are set to 1e-6, the convergence residual values of H2O
Synthesize the above: according to the inlet and outlet water tem­
are set to 1e-5, and the convergence residual values of other equations
perature of the heat exchanger tube of the cooling tower, the cooling
are all set to: 1e-4.
efficiency of the cooling tower is calculated by Eq. (19).
tw,in − tw,out
Table 2 η= × 100% (19)
Boundary conditions. tw,in − twb
Boundary condition Parameters
Where: η represents cooling efficiency, twb represents wet bulb
Time Steady state temperature.
Inlet wind velocity 2–4 m/s
Inlet air temperature 25–37 ◦ C
Outlet pressure 0 3.6. Mesh independency
Gravity acceleration − 9.81
Spray density 0.01–0.08 Grid quality is the premise to ensure the accuracy of calculation. In
Spray water temperature 35–41 ◦ C this study, ICEM is selected to divide the geometric model into struc­
Turbulence intensity 10%
tured grids, and the height of the first layer grid is determined to be

4
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

0.001. The quality and quantity of the wall grid have a great influence the increase of the inlet wind velocity. However, the calculated results of
on the heat exchange. in order to accurately capture the heat and mass the DRW model are more close to the experimental results, and the error
transfer between air and spray water, the wall grid is refined. At the between them is less than 3.15%. Synthesize the above: determine the
same time, the quantity and quality of the grid directly affect the ac­ combined model of k-ε realizable and DRW as the calculation model in
curacy and time of the Fluent calculation results, so it is necessary to numerical research.
verify the irrelevance of the grid.
In this study, the number of grids in the calculation domain is divided 4. Results and discussions
into five different orders of magnitude: 24,153, 47,369, 81,841, 155,305
and 306,741, respectively. The temperature of the air outlet is selected 4.1. Flow field characteristics
as the measurement index to study the influence of the number of grids
on the numerical simulation results to ensure the accuracy of the 4.1.1. Analysis of turbulent kinetic energy (Reviewer 2, Comment 7)
calculation results. When the number of grids increases from 24,153 to The distribution of turbulent kinetic energy in the cooling tower can
155,305, the air temperature of the air outlet of the cooling tower also directly reflect the turbulence intensity in the tower. In this simulation
increases, but when the number of grids exceeds 155,305, the air outlet calculation condition, the wind speed is set to 3 conditions, which are 2
temperature of the cooling tower has no further rising trend. Therefore, m/s ~4 m/s respectively. The spray density is 0.03 kg m− 2•s− 1, the
considering the above, the grid of the order of 155,305 is selected in the spray water temperature is 306 K, and the inlet air temperature is set at
numerical calculation. 302 K. The distribution of turbulent kinetic energy in the tower at
different wind speeds is shown in Fig. 3.
As can be seen from Fig. 3, when the air enters the tower, the air
3.7. CFD validation resistance of the circular tube becomes larger and the turbulent kinetic
energy of the circular coil is larger than that of the elliptical tube because
On the basis of determining the best order of magnitude grid, the of its larger upwind area. Because the new elliptical finned tube has the
turbulence used needs to be verified. In this study, three typical k-ε effect of disturbing the air, the air resistance increases, the turbulent
turbulence models (Standard, RNG and Realizable) are selected for kinetic energy in the tower is enhanced, and the turbulence is more
simulation calculation, and the air temperature at the air outlet of the intense than that of the elliptical tube. With the increase of the flow
cooling tower is also selected as the evaluation parameter index. velocity in the tower, the turbulence of the new finned tube becomes
Compared with the experimental results, the best calculation model is more and more intense, which further shows that the disturbance of the
obtained. The specific settings of numerical calculation are as follows: fin leads to the violent fluctuation of the velocity on both sides of the
the velocity range of inlet air is 2 m/s ~4 m/s, the mass flow rate of tube wall, and the greater the inlet wind velocity, the greater the tur­
spray water is 0.025 kg m− 2 s− 1, and the temperature of inlet air is 35 ◦ C. bulent kinetic energy on both sides of the tube wall.
As shown in Fig. 2, when the inlet wind velocity increases, the exit air
temperature of the three turbulence models increases, but the change 4.1.2. Characteristic analysis of pressure field
ratio of the three models is not the same. The difference between the The kinetic energy loss in the tower is an important index to evaluate
outlet air temperature of the Realizable model and the experimental the performance of the cooling tower. In the simulation calculation, the
data is only 5.24%, while the errors between the other two models RNG inlet wind velocity is set to three working conditions: 2 m/s, 3 m/s and 4
and Standard and the experimental data are 21.46% and 26.78%, m/s. The spray density is 0.03 kg m− 2 s− 1, the spray water temperature
respectively. And the influence of turbulent viscosity is added to the is 306 K, and the inlet air temperature is 302 K. The pressure distribution
Realizable model, so the k-ε Realizable turbulence model is selected as in the tower under different wind velocity is shown in Fig. 4.
the numerical simulation. As shown in Fig. 4, the overall pressure loss of the cooling tower
While determining the k-ε Realizable model, the effects of DRW usually shows a linear trend with the increase of inlet wind velocity,
(random walk model) and without DRW model on the thermal perfor­ which inevitably leads to a certain loss of kinetic energy when air flows
mance of the cooling tower are verified again. As can be seen in Fig. 2, in the tower, that is, the increase of pressure drop in the tower. For the
the calculated results of the two models have the same trend as the elliptical heat exchanger tube, the air flow velocity in the tower is fast,
experimental results, and the temperature of the outlet air increases with the resistance along the way is small, and the loss of air kinetic energy in
the tower is small; for the new elliptical finned tube, the air flow resis­
tance in the tower is increased because of the fins added on the tube
wall, that is, the air inertia and flow resistance in the tower are
increased. But it also increases the contact area between the pipe wall
and spray water to make the heat transfer between them more sufficient.
The heat exchanger tubes of the three structures produce greater pres­
sure at the air inlet, and the pressure at the outlet of the air flow is
relatively small, and the kinetic energy loss of the new elliptical finned
tube is 1.09 and 1.15 times higher than that of the other two kinds of
heat exchanger tubes.

4.1.3. Characteristic analysis of temperature field

The temperature field distribution in the tower directly illustrates the
heat transfer between the fluids in the cooling tower. In this simulation,
the spray density is set to three working conditions, which are 0.02 kg
m− 2 s− 1, 0.05 kg m− 2 s− 1 and 0.07 kg m− 2 s− 1, inlet air velocity is 3 m/s,
spray water temperature is 306 K, inlet air temperature is 302 K. The
temperature field distribution in the tower under different spray density
is shown in Fig. 5.

Fig. 2. Turbulence model validation.

5
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Fig. 3. Distribution of turbulent kinetic energy in tower at different wind velocity.

Fig. 4. Pressure distribution in the tower at different velocity.

4.2. Effect of boundary conditions on the performance of the density is 0.05 kg m− 2•s− 1, and the spray water temperature is 308 K.
cooling tower (Reviewer 2, Comment 11) The thermal characteristics of the closed wet cooling tower at different
inlet wind velocities are shown in Fig. 6.
4.2.1. Comparison of the effect of wind velocity on cooling efficiency As can be seen from Fig. 6, the cooling efficiency increases with the
According to the actual project, in the simulation study, the inlet increase of the inlet wind velocity, when inlet wind velocity is 2 m/s, the
wind velocity is set to 2–4 m/s, the air temperature is 306 K, the spray cooling efficiency is 34.06%, and when inlet wind velocity increases to

6
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Fig. 5. Distribution of temperature field in tower under different spray density.

Fig. 6. Influence of different inlet wind velocity on simulation results. Fig. 7. Influence of different spray densities on simulation results.

4 m/s, the cooling efficiency increases by 8.63%. When inlet wind ve­ density, the cooling efficiency first increases with the spray density and
locity increases from 2 m/s to 4 m/s, the heat transfer coefficient does then tends to stabilize. When the spray density is 0.02 kg m− 2•s− 1, the
not change with the increase of wind velocity. This is due to the fact that cooling efficiency is 31.47%, and when the spray density increases to
with the increase of inlet wind velocity, the residence time of the air in 0.07 kg m− 2•s− 1, the cooling efficiency increases by 8.75%. When the
the tower becomes shorter, which strengthens the convective heat spray density increases from 0.02 kg m− 2•s− 1 to 0.07 kg m− 2•s− 1, the
transfer between water and gas, which lowers the water temperature at heat transfer coefficient increases from 915.1671 W m− 2•◦ C− 1 to
the outlet of the heat exchanger coil and improves the cooling efficiency. 923.5659 W m− 2•◦ C− 1, an increase of about 6.43%. The reason is that
with the continuous increase of spray density, the distribution of the
4.2.2. Comparison of effects of spray density on cooling efficiency liquid film on the pipe wall is more uniform, and the heat transfer be­
In this study, according to the actual working conditions, the range of tween the spray water and the pipe wall is more sufficient. When the
spray density is 0.02–0.07 kg m− 2•s− 1, the inlet wind velocity is 3 m/s, wind velocity is constant, the cooling efficiency and heat transfer coef­
the air temperature is 306 K, and the spray water temperature is 308 K. ficient will increase.
The thermal characteristics of closed wet cooling towers with different
spray densities are shown in Fig. 7. 4.2.3. Comparison of effects of wet bulb temperature on cooling efficiency
As can be seen in Fig. 7, with the continuous increase of spray According to the air wet bulb temperature in different seasons, the

7
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

range of wet bulb temperature is 278–308 K, the inlet wind velocity is 3 high, although the heat exchange between the water film outside the
m/s, the spray water density is 0.05 kg m− 2•s− 1, and the spray water tube and the circulating process water inside the tube is increased, the
temperature is 308 K. The changes in the thermal characteristics of the liquid film on the tube wall is thicker. it not only affects the heat transfer
cooling tower under different wet bulb temperatures are shown in Fig. 8. performance, but also increases the evaporation loss of spray water; the
As can be seen from Fig. 8, as the temperature of the air wet bulb smaller spray density makes the pipe wall can not be covered by water
temperature is increasing, the cooling efficiency is constantly film, and the heat exchange of the cooling tower is insufficient. The
decreasing; when the air wet bulb temperature is 278 K, the cooling specific parameters of this numerical study are as follows: the inlet wind
efficiency is 37.45%, when the air wet bulb temperature increases to velocity is 3 m/s, the inlet air temperature is 303 K, the spray water
308 K, the cooling efficiency is reduced by 5.89%. When the air wet bulb density is 0.02–0.07 kg m− 2 s− 1, and the spray water temperature is 305
temperature rises from 278 k to 308 K, the heat transfer coefficient is K. Fig. 10 shows the changing trend of thermal performance of closed
reduced from 921.8145 W m− 2•◦ C− 1 to 906.5671 W m− 2•◦ C− 1, and wet cooling tower under different spray density.
reduced by 1.68%. This is because the cooling efficiency of the cooling With the increase of spray water density, the thickness of liquid film
tower is influenced by the difference between the temperature of the on the surface of pipe wall also increases, resulting in greater heat
circulation process and the temperature of the inlet wet bulb, the larger transfer resistance and slower increase of cooling efficiency. Moreover,
the difference, and the cooling efficiency will continue to increase. the cooling efficiency of the new heat exchanger tube with higher
cooling efficiency is 1.27 and 1.17 times higher than that of the other
two heat exchanger tubes, which coincides with the changing trend of
4.3. Effect of air velocity
heat transfer coefficients.
The degree of wetting on the surface of the pipe wall is constantly
Wind velocity is an important operation parameter that affects the
changing, and when the thickness of the water film reaches a certain
thermal performance of closed wet cooling tower. The cooling efficiency
degree, the thermal resistance will also become higher, so that the heat
can be changed by adjusting the wind velocity in the tower, but the
exchange between the circulating water in the tube and the spray water
continuous increase of wind velocity means that the air resistance and
outside the pipe becomes less, and the increase of heat transfer coeffi­
fan energy consumption in the tower are also increasing. Therefore, it is
cient becomes more stable, which also shows that the new heat transfer
particularly important to determine a reasonable wind velocity for the
tube can significantly enhance the turbulence, enhance the convective
thermal performance of the cooling tower. The specific parameters of
heat transfer between air and spray water.
this numerical study are set as follows: the range of wind velocity is 2 ~
4 m/s, the inlet air temperature is 303 K, the spray water flow rate is
4.5. Effect of air temperature
0.05 kg m− 2 s− 1, the spray water temperature is 305 K, and the spray
height is set to 20 mm. Fig. 9 shows the changing trend of thermal
The inlet air wet bulb temperature varies with the season, which is an
performance in the tower under different inlet wind velocity.
important environmental factor affecting the cooling efficiency and heat
The increase of air velocity in the tower increases the rate of heat
transfer coefficient of the closed wet cooling tower. The specific pa­
transfer between spray water and air. In addition, the increase of air
rameters of the simulation calculation are as follows: the inlet wind
velocity in the tower helps to reduce the temperature difference between
velocity is 3 m/s, the inlet air temperature is 278 K, 287 K, 293 K, 300 K,
the outlet water temperature of the circulating process water in the tube
308 K, the mass flow rate of spray water is 0.05 kg m− 2 s− 1, and the
and the wet bulb temperature of the inlet air. Generally speaking, the
spray water temperature is 305 K. Fig. 11 shows the changing trend of
increase of inlet wind velocity has a positive effect on the cooling effi­
the thermal performance of the cooling tower under different inlet air
ciency, heat transfer coefficient of closed wet cooling tower. However,
wet bulb temperatures.
increasing the intake velocity will inevitably increase the pressure drop
The cooling efficiency is affected by the temperature difference be­
of the closed wet cooling tower and the energy consumption of the fan.
tween the circulating process water in the tube and the wet bulb tem­
perature of the inlet air, the greater the difference is, the stronger the
4.4. Effect of spray density cooling efficiency of the cooling tower is; with the change of the season,
the moisture content of the air and the wet bulb temperature will change
The spray density has an important influence on the heat transfer accordingly, when the moisture content of the air is low, the heat ex­
performance of the closed wet cooling tower. When the spray density is change capacity between the air and the spray water is stronger; on the
contrary, it will cause the phenomenon of poor cooling efficiency.
With the increase of air wet bulb temperature, the temperature dif­
ference between water film and air decreases, and the heat and mass
exchange between water and air decreases. In this regard, the higher wet
bulb temperature is not conducive to the heat exchange in the cooling
tower. In addition, the humidity of the gas-liquid interface and the
temperature of the water film remain basically unchanged, which leads
to a downward trend in the heat transfer coefficient between the water
film and the air in the tower.

4.6. Effect of spray water temperature

In order to explore the effect of spray water temperature on the heat

transfer performance of closed wet cooling tower, when the spray
temperature is higher, the evaporation of spray water in the tower will
become larger, resulting in more evaporation loss, so determining a
reasonable spray water temperature is an important work. The specific
parameters of the simulation calculation are as follows: the temperature
range of spray water is between 10 ◦ C and 35 ◦ C, the mass flow rate of
spray water is 0.05 kg m− 2 s− 1, the inlet air velocity is 3 m/s, the inlet air
Fig. 8. Influence of different wet bulb temperatures on simulation results. temperature is 307 K, and the air humidity is 64%. Fig. 12 shows the

8
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Fig. 9. Transformation trend of thermal performance in tower under different inlet wind velocity.

Fig. 10. Transformation trend of thermal performance in tower under different spray density.

Fig. 11. Variation curve of thermal performance of cooling tower under different wet bulb temperature.

changing trend of thermal performance in the tower at different spray circular heat transfer tube, when the spray water temperature increases
temperatures. from 10 ◦ C to 35 ◦ C, the heat transfer coefficient increases by 9.78
As can be seen from Fig. 12, when the spray water temperature of the w⋅m− 2⋅s− 1. When the spray water temperature of the new heat transfer
elliptical heat transfer tube increases from 10 ◦ C to 35 ◦ C, the heat tube increased from 10 ◦ C to 35 ◦ C, the heat transfer coefficient increases
transfer coefficient increases by 9.45 w⋅m− 2⋅s− 1. For the ordinary by 10.56 w⋅m− 2⋅s− 1, thus it can be seen that the heat transfer coefficients

9
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

Fig. 12. Changed trend of thermal performance in the tower under different spraying temperatures.

of the three kinds of heat exchanger tubes with different structures in­ respectively. When the spray density increased from 0.02 kg m− 2
crease by 27%, 32% and 36%, respectively. This is because the increase s− 1 to 0.05 kg m− 2 s− 1, the cooling efficiency increased by 4.31%,
in spray temperature causes more water to evaporate and enhances the 4.03% and 4.83%, respectively, and the heat transfer coefficient
effect of convective heat transfer. Respectively, so the heat transfer co­ increased by 23%, 26% and 30%, respectively. When the air wet
efficient and the thermal performance of the cooling tower will increase bulb temperature increased from 5 ◦ C to 35 ◦ C, the cooling effi­
with the increase of spray water temperature. ciency decreased by 6.62%, 6.42% and 6.51%, respectively. And
the heat transfer coefficient decreased by 2.5%, 1.7% and 1.9%,
5. Conclusions respectively.
(5) The new elliptical finned tube keeps the advantages of the ellip­
The effects of heat exchanger tubes with different structures on the tical heat exchanger tube, and its high radius of curvature can
distribution characteristics of flow field (velocity field, temperature accelerate the speed of air flow in the tower, thus increasing the
field, pressure field) in the tower are simulated by using Fluent nu­ Reynolds number. The fins of the new elliptical finned tube will
merical calculation software, and the effects of operation parameters aggravate the air disturbance on both sides of the tube wall and
such as wind velocity and spray density on the thermal performance of further increase its heat transfer coefficient and Reynolds num­
cooling tower are analyzed. The specific conclusions are as follows: ber. Under the same conditions, the new elliptical finned tube has
higher heat transfer coefficient and mass transfer coefficient,
(1) Velocity field distribution characteristics: with the continuous which can greatly improve the cooling efficiency. (Reviewer 1,
increase of air mass flow, the distribution of the whole velocity Comment 1)
field is more uniform, and there is an obvious cylindrical flow on
both sides of the pipe wall, especially when the air velocity is Declaration of competing interest
high. The phenomenon of air flow around the wall of the new
elliptical finned tube is particularly prominent. The flow velocity The authors declare that they have no known competing financial
on both sides of the tube wall is obviously higher than that on interests or personal relationships that could have appeared to influence
both sides of the other heat exchanger tubes. That is, when the air the work reported in this paper.
flow rate in the tower is 3.5 kg m− 2 s− 1, the wind velocity of the
new heat exchanger tube is the highest on both sides. Acknowledgement
(2) Pressure field distribution characteristics: when the inlet air ve­
locity increases from 2 m/s to 4 m/s, the heat exchanger tubes of This research is funded by Changzhou Key Science and Technology
the three structures produce greater pressure at the air inlet, and Achievements Conversion and Industrialization Plan (Grant No.
the pressure at the outlet of the air flow is relatively small, while CE2017125), Major Projects of Natural Science Research of the Jiangsu
the loss in the new tube tower is 1.09 and 1.15 times that of the Higher Education Institutions (Grant No.18KJA480002), Key R&D
other two structures. Project (Technology Support for Social Development) of Changzhou City
(3) Temperature field distribution characteristics: the outlet tem­ (Grant No.CE20195035), Jiangsu Key Laboratory of Green Process
perature of the heat exchanger tube with three different struc­ Equipment (Grant No.GPE201805), Changzhou Key Science and Tech­
tures increases with the increase of spray density. For new nology Achievements Conversion and Industrialization Plan (Grant No.
elliptical finned tube, the air forms a strong eddy current on both CE2020043) and Key R & D Plan of Yangzhou City (Industry Prospect
sides of the tube wall, which strengthens the heat transfer be­ and Common Key Technology) (Grant No.YZ2020005). They are
tween the air and the high temperature spray water. The outlet gratefully acknowledged.
temperature is higher and the cooling effect is better, which is
1.65 and 2.56 times higher than that of the other two heat
References
exchanger tubes.
(4) The influence of different factors on the thermal performance of [1] R.A. Kheneslu, A. Jahangiri, M. Ameri, Interaction effects of natural draft dry
the cooling tower: when the wind velocity increases from 2 m/s cooling tower (NDDCT) performance and 4E (energy, exergy, economic and
environmental) analysis of steam power plant under different climatic conditions,
to 5 m/s, the cooling efficiency of the heat exchanger tube of the
Sustainable Energy Technologies and Assessments 37 (2020).
three structures is 8.67%, 9.02% and 9.13% respectively, and the [2] S.Y. Yoo, J.H. Kim, K.H. Han, et al., Thermal performance analysis of heat
heat transfer coefficient increases by 32%, 35% and 37%, exchanger for closed wet cooling tower using heat and mass transfer analogy,
J. Mech. Sci. Technol. 24 (2010) 893–898.

10
R. Zhao et al. International Journal of Thermal Sciences 168 (2021) 107064

[3] P. Song, B. Wang, X. Li, et al., Experimental research on heat and mass transfer [12] W.Y. Zheng, D.S. Zhu, G.Y. Zhou, et al., Thermal performance analysis of closed
characteristics of cross-flow closed-type heat-source tower, Appl. Therm. Eng. 135 wet cooling towers under both unsaturated and supersaturated conditions, Int. J.
(2018) 289–303. Heat Mass Tran. 55 (26) (2012) 7803–7811.
[4] J.J. Jiang, X.H. Liu, Y. Jiang, Experimental and numerical analysis of a cross-flow [13] Y. Zhou, X. Zhu, X. Ding, et al., Theoretical investigation on thermal performance
closed wet cooling tower, Appl. Therm. Eng. 61 (2013) 678–689. of new structure closed wet cooling tower, Heat Tran. Eng. 51 (39) (2018)
[5] T.A. Tahseen, M. Ishak, M.M. Rahman, An overview on thermal and fluid flow 2079–2086.
characteristics in a plain plate finned and un-finned tube banks heat exchanger, [14] C.X. Xiao, H. Chang, X. Tao, et al., Deciphering the thermal and hydraulic
Renew. Sustain. Energy Rev. 43 (2015) 363–380. performances of closed wet cooling towers with plain, oval and longitudinal fin
[6] S.M. Qasim, M.J. Hayder, Parametric study of closed wet cooling tower thermal tubes, Appl. Therm. Eng. (120) (2017) 203–218.
performance, Mater. Sci. Eng. 7 (2017) 23–28. [15] M.K. Mansour, M.A. Hassab, Novel approach of thermal modeling of partially dry-
[7] J. Sutherland, Analysis of mechanical-draught counter-flow air/water cooling wet chilled water cooling tube under unit and nonunit Lewis number conditions,
towers, J. Heat Tran. 105 (3) (1983) 576. Numer. Heat Tran., Part B: Fundamentals 8 (52) (2016) 1–18.
[8] C.Q. Ren, Corrections to the simple effectiveness-NTU method for counter-flow [16] E. Masi, J. Bellan, K.G. Harstad, et al., Multi-species turbulent mixing under
cooling towers and packed bed liquid desiccant–air contact systems, Int. J. Heat supercritical-pressure conditions direct numerical simulation and analysis
Mass Tran. 51 (2) (2008) 237–245. revealing species decomposition, J. Fluid Mech. 7 (21) (2013) 578–626.
[9] H.J. Cui, N.P. Li, J.Q. Peng, et al., Study on the dynamic and thermal performances [17] X. Li, H. Gurgenci, Z. Guan, et al., A review of the crosswind effect on the natural
of a reversibly used cooling tower with upward spraying, Energy 44 (96) (2016) draft cooling towers, Appl. Therm. Eng. (150) (2019) 250–270.
268–277. [18] ANSYS, ANSYS FLUENT 12.0 Theory Guide, 2009.
[10] J.A. Heyns, D.G. Kröger, Experimental investigation into the thermal-flow [19] ANSYS, ANSYS FLUENT 12.1 User’ Guide, 2009.
performance characteristics of an evaporative cooler, Appl. Therm. Eng. (5) (2010) [20] S.A. Zhao, Cooling Tower Process Principle, China Construction Industry Press,
492–498. 2015 (in Chinese).
[11] K. Singh, R. Das, A feedback model to predict parameters for controlling the
performance of a mechanical draft cooling tower, Appl. Therm. Eng. (S1) (2016)
519–530.

11